Research And Implementation Of Software Testing Method Based On Feedback Control

In the software testing process and testing requirements of users often change, this time testing personnel should as far as possible to meet user needs. At this time, if we restart all the testing process, so the cost of testing work and all the test work done before will become meaningless. And the testing results for the testing process has very important reference value. If in this test, the execution of test cases which have the strong ability to found defects is preferred, then it can improve the testing efficiency to a certain extent. Therefore, through the idea of feedback, if the test results as a part of the testing strategy in test process which the user’s demands change, and the demand change as a part of the decision at the same time, it will greatly improve the test efficiency, and meet the needs of users as soon as possible, this is a closed-loop feedback process.When the user’s requests change, then we need to through the changing requests to make a standard for the testing process and design the testing to improve testing efficiency, especially the requests for hardware and software. Cloud computing environment can provide flexible and extensible software and hardware resources for different applications, so whether developers or testers can avoid the cumbersome purchase of hardware resources, reduce the cost.This paper first through the description of the improved feedback test model based on the feedback control. Then it introduces the various parts of the improved feedback model, including:the test requirement definition principle and the specific content of the test strategy. Finally introduces the deployment process for the testing environment. And in the feedback process, the most important part is the testing strategy, it contains the test environment build, test case selection and the testing needs to determine the standard stop test and the constraints. This paper proposes some improvements based on the original Markov decision model and constructs a Markov decision model, which is based on the user’s detected defects rate as the test criterion and the testing cost as the constraint. Then according to the fault rate, we put forward a test case selection strategy based on the fault rate, which can select the test case with the highest fault rate, and the test cases are grouped according to the fault rate, this result can be used to carry out statistical analysis of the test results. Finally, a new software testing method is proposed based on the improved Markov decision model and the test case selection strategy based on the fault rate. This method is a specific implementation of the test strategy in the feedback model. According to some comparative simulation experiments, it is proved that the software test method based on feedback in this paper is effective.